Environ. Sci. Technol. 2002, 36, 4990-4997
Accumulation of COGEMA-La Hague-derived Reprocessing Wastes in French Salt Marsh Sediments A N D R E W B . C U N D Y , * ,† IAN W. CROUDACE,‡ PHILLIP E. WARWICK,‡ JUNG-SUK OH,‡ AND SIMON K. HASLETT§ School of Chemistry, Physics and Environmental Science, University of Sussex, Brighton, BN1 9QJ, U.K., Southampton Oceanography Centre, Southampton, SO14 3ZH, U.K., and School of Science and the Environment, Bath Spa University College, Bath, BA2 9BN, U.K.
Over the past five decades, authorized low-level discharges from coastal nuclear facilities have released significant quantities of artificial radionuclides into the marine environment. In northwest Europe, the majority of the total discharge has derived from nuclear reprocessing activities at Sellafield in the United Kingdom and COGEMALa Hague in France. At the Sellafield site, a significant amount of the discharges has been trapped in offshore fine sediment deposits, and notably in local coastal and estuarine sediments, and much research has been focused on understanding the distribution, accumulation, and reworking of long-lived radionuclides in these deposits. In contrast, there are few high-resolution published data on the vertical distribution of radionuclides in fine-grained estuarine sediments near, and downstream of, COGEMALa Hague. This paper therefore examines the vertical distribution of a range of anthropogenic radionuclides in dated salt marsh cores from two estuaries, one adjacent to, and the other downstream of, the COGEMA-La Hague discharge point (the Havre de Carteret at Barneville-Carteret and the Baie de Somme, respectively). The radionuclides examined show a vertical distribution which predominantly reflects variations in input from COGEMA-La Hague (albeit much more clearly at Barneville-Carteret than at the Baie de Somme site), and Pu isotopic ratios are consistent with a La Hague, rather than weapons’ fallout, source. Because of sediment mixing, the marshes apparently retain an integrated record of the La Hague discharges, rather than an exact reproduction of the discharge history. Sorption of radionuclides increases in the order 90Sr < 137Cs < 60Co < 239,240Pu, which is consistent with Kd values reported in the literature. In general, the radionuclide activities observed at the sites studied are low (particularly in comparison with salt marsh sediments near the Sellafield facility), but are similar to those found in areas of fine sedimentation in the central Channel. These marshes are not major sinks for discharged reprocessing wastes.
* Corresponding author phone: +44 1273 678737; fax: +44 1273 677196; e-mail:
[email protected]. † University of Sussex. ‡ Southampton Oceanography Centre. § Bath Spa University College. 4990
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 23, 2002
Introduction Over the past five decades, authorized low-level discharges from coastal nuclear facilities have released significant quantities of artificial radionuclides into the marine environment. In northwest Europe, the majority of the total discharge has derived from nuclear reprocessing activities at Sellafield in the U.K. and COGEMA-La Hague in France. Discharges commenced in 1952 and 1967, respectively, for the two plants. Using 90Sr as an example, COGEMA-La Hague released an estimated 1600 TBq 90Sr (or ca. 300 g, value decaycorrected to 1995) over the period 1967-1995 (1), while Sellafield released 6000 TBq (ca. 1200 g, value decay-corrected to 1995) over the same period. The radionuclides discharged have been widely dispersed, reaching Scandinavia, the Arctic, and the eastern seaboard of the United States (2, 3). A considerable volume of research has been published on the behavior of radionuclides following their release from Sellafield (e.g., 4-6) and COGEMA-La Hague (e.g., 1, 7, 8). Whereas more soluble radionuclides (e.g., 125Sb (8) and 99Tc (9)) tend to disperse along water transport pathways, particlereactive radionuclides (e.g., Pu, 241Am) may interact with fine, clay-rich suspended sediments and subsequently be deposited and buried in sub- and inter-tidal mud deposits and coastal salt marshes. These sedimentary environments may therefore act as important short- to medium-term sinks for radionuclides, building up a significant inventory of artificial radionuclides and forming a potential pathway for local population exposure, as has been observed around the Sellafield site (10-13). Significant amounts of the historical discharges from Sellafield have been trapped in offshore fine sediment deposits (the Sellafield “mud-patch”), and notably in local coastal and estuarine sediments, and much research has focused on understanding the distribution, accumulation, and reworking of long-lived radionuclides in these deposits. In contrast, despite the presence of large areas of mudflat and saltmarsh in the estuaries of the north French coast, there are few high-resolution published data on the vertical distribution of radionuclides in fine-grained estuarine sediments near, and downstream of, COGEMA-La Hague. The importance of these estuarine sediments as a medium-term sink for COGEMA-La Hague-derived radionuclides remains uncertain. This paper addresses this by examining the vertical distribution of a range of anthropogenic radionuclides in dated sediment cores from two estuaries: one adjacent to, and the other downstream of, the COGEMA-La Hague pipeline (the Havre de Carteret at Barneville-Carteret and the Baie de Somme, respectively). The extent to which the marshes retain a record of COGEMA-La Hague discharges is examined, and their importance as sinks for different radionuclides is assessed.
Methods Salt marsh cores were collected using an Eijkelkamp gouge (Barneville-Carteret) or a hand-driven PVC coring tube (Baie de Somme). Sediment compaction caused by the coring procedure was measured and was found to be negligible in all cases. On return of the cores to the laboratory, they were sliced open and described before being cut into 1- or 2-cm depth increments for further analysis. 210Pb activity was determined by a proxy method through alpha spectrometric measurement of its granddaughter nuclide 210Po. The method employed was based on that of Flynn (14), using double acid leaching of the sediment with 209Po as an isotopic tracer and autodeposition of the Po isotopes in the leachate on to silver disks. Detection limits were 1 Bq/kg. 137Cs, 60Co, and 241Am 10.1021/es020098c CCC: $22.00
2002 American Chemical Society Published on Web 11/01/2002
FIGURE 1. Northern France and the central part of the English Channel/La Manche. The locations of the COGEMA-La Hague reprocessing facility and the AEA Winfrith nuclear facility are marked with open circles. The study areas (Barneville-Carteret and Baie de Somme) are marked with closed circles.
activities were determined by gamma spectrometry with Canberra HPGe well-type detectors. Uncertainties were typically in the order of 4% (1σ) and detection limits were 1 Bq/kg. At Barneville, 241Am was further constrained by alpha spectrometric measurements (see below). Sediment aliquots of 5 g were spiked with 243Am, 242Pu, and 85Sr. The sediments were leached twice with aqua regia (75% HCl/25% HNO3). The leachate was evaporated to dryness, and the residue was dissolved in nitric acid. Oxalic acid was added to the solution, and the pH was adjusted to 4.5 resulting in the precipitation of calcium oxalate. The precipitate was isolated by filtration and ignited overnight at 450 °C. The resulting residue was dissolved in a minimum amount of 8 M nitric acid. Iron hydroxide was precipitated through the addition of ammonia solution and the supernatant was reserved for 90Sr analysis. The Fe(OH)3 precipitate was dissolved in nitric acid + ascorbic acid, and the Pu and Am were isolated using a combination of anion exchange and extraction chromatographic techniques (15), prior to alpha spectrometric measurement. 239Pu and 240Pu were also determined on four samples from BarnevilleCarteret via multicollector ICP-MS, using the method of Taylor et al. (16). The supernatant from the Fe(OH)3 precipitation was taken for 90Sr analysis. The supernatant was evaporated to dryness and the residue dissolved in 8M HNO3. The Sr was isolated on an Eichrom Sr resin column and eluted with 10 mL of water directly into the scintillation vial. The vial was counted by gamma spectrometry to determine the recovery of 85Sr. The 90Sr activity was then counted repeatedly over a period of 2 weeks by Cerenkov counting (using a Wallac 1220 Quantulus low-level liquid scintillation counter) to monitor the ingrowth of the 90Y daughter. The 90Sr activity of the sample was calculated from the 90Y daughter activity (17). All samples were analyzed using a Philips PW1400 X-ray fluorescence spectrometer to obtain compositional data.
TABLE 1. Major Sources and Inputs of Artificial Radionuclides to the Channel; Data from Boust (1)a
radionuclide
period of peak discharge from COGEMALa Hague
137Cs
1971
654 TBq
Inputs from atmospheric weapons testing and reactor accidents (via direct fallout and inwash from the Atlantic) ) 800 TBq. Minor inputs from the AEA Winfrith and Sellafield facilities, U.K.?
90Sr
1982-1985
1600 TBq
Inputs from atmospheric weapons testing and reactor accidents (via direct fallout and inwash from the Atlantic) ) 550 TBq (estimated assuming 137Cs/90Sr ratio of 1.45 in weapons fallout, 19).
60Co
1983-1985
33 TBq
Input from AEA Winfrith, U.K. ) 20 TBq. Prevailing water and sediment transport pathways mean that little of this is likely to reach the French coastline.
238Pu
1974?
1.3 TBq (estimated)
Inputs from atmospheric weapons testing (via direct fallout and inwash from the Atlantic) ) 0.3 TBq. Minor inputs from the AEA Winfrith and Sellafield facilities, U.K.?
239,240Pu
1974
2.5 TBq (estimated)
Inputs from atmospheric weapons testing (via direct fallout and inwash from the Atlantic) ) 8 TBq. Minor inputs from the AEA Winfrith and Sellafield facilities, U.K.?
241Am
1974?
?
Inputs from atmospheric weapons testing (via direct fallout and inwash from the Atlantic). Minor inputs from the AEA Winfrith and Sellafield facilities, U.K.?
a
TBq
total discharge from La Hague since 1967 (cumulative to 1995, decay-corrected)
other main sources
For comparison: over the period 1967-1995 the Sellafield facility discharged approximately 26 000 TBq 137Cs, 3600 TBq 90Sr, 5 TBq 60Co, 100 238 Pu, 600 TBq 239,240Pu, and 500 TBq 241Am (activities decay-corrected to 1995).
VOL. 36, NO. 23, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4991
Samples were pelletized for trace element determinations and fused for major element determinations. The methods used are well-established and are detailed in Croudace and Williams-Thorpe (18). Precision is nominally 1% rsd for major elements and 5% rsd for trace elements. Accuracy was assessed by comparing a range of reference sample determinations (e.g., USGS MAG-1) with recommended values and was generally within 10% of the working values quoted in the literature. Sample mineralogy was determined using a Philips X-ray diffractometer. COGEMA-La Hague Facility and Other Sources of Artificial Radioactivity. COGEMA-La Hague and BNFL Sellafield are the two major nuclear fuel reprocessing facilities in northern Europe. The COGEMA-La Hague facility (Figure 1) was developed in the 1960s, and nuclear reprocessing commenced on January1, 1967. Historically, the facility has discharged a range of radionuclides (including 90Sr, 125Sb, 60 Co, Pu isotopes, 241Am, and 137Cs) under authorization into The Channel and into the atmosphere, and is currently (and historically for many radionuclides) the most significant source of artificial radioactivity into Channel waters (Table 1). Other significant historical sources of radioactivity to the study area (Table 1) are atmospheric nuclear weapons testing and the AEA Winfrith nuclear facility on the southern U.K. coast (Figure 1). Global dispersion of radionuclides from atmospheric nuclear weapons testing occurred from 1954 onward, with the most significant fallout occurring between the 1950s and 1970s. Peak fallout occurred in 1963, following increased nuclear testing activity in the early 1960s. The AEA Winfrith site began discharging radioactive effluent in 1970. This effluent mainly consisted of the major activation products of Fe, Co, Cr, Mn, Ni, and Zn released following annual cleaning of the SGHW reactor at the site, although minor quantities of fission products and actinides were also released following fuel rod inspection work (20). Discharges peaked during the early 1980s as the operations at the site increased. In 1990, following the shut-down and subsequent decommissioning of the SGHW reactor, discharged radionuclide activities fell dramatically. The AEA Winfrith discharges were predominantly swept along the U.K. coastline toward the North Sea along prevailing sediment and water transport pathways (21), and while some limited crossChannel movement may have occurred, the comparatively low activities discharged, coupled with mixing and dispersion processes, mean that little significant AEA Winfrith-derived radioactivity is likely to be found at the sites examined here (Table 1). Discharges from BNFL Sellafield have also been widely dispersed in northwestern European waters. Peak discharges of fission products and actinides from the Sellafield facility occurred in the mid-1970s, and a small fraction of these discharges may have entered the Channel several years later. The influence of the BNFL Sellafield discharges, however, is insignificant in the central Channel compared to the other sources discussed above (22). Hydrography. The coastal area into which COGEMA-La Hague discharges radionuclides is characterized by high tidal current velocities of ca. 2.5 m/s (Figure 2a). This means that, theoretically, these radionuclides should be rapidly dispersed. Most of the discharge is swept eastward toward the North Sea, following general water and sediment transport pathways which move west to east along the northern French coastline (8). A fraction of the discharge (approximately 5%) is also swept southwest toward the Channel Islands on a local current gyre (24). Typically, the highest activities of discharged radionuclides are observed between Barneville-Carteret and the Pointe de Barfleur (Figure 1; 25, 26), with a slow decline in activity with distance eastward. Guegueniat et al. (8) identified two distinct eastward transport pathways: a rapid central Channel route and a slow coastal route (Figure 2b). These authors note transit times for dissolved radionuclides 4992
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 23, 2002
FIGURE 2. A. Peak, near-surface tidal current speeds in the central and eastern Channel (m/s), after 1 and 23. B. Schematic map of transport pathways for dissolved radionuclides discharged from COGEMA-La Hague (after 8 and 24). See text for discussion. for each route (i.e., the time taken for a pulse of discharged radionuclides to pass into the North Sea) of approximately 3-4 months and 7-8 months, respectively. As noted by Boust (1), however, the transit times for particle-associated radionuclides may be substantially longer than that for dissolved radionuclides. Study Areas: Barneville-Carteret and Baie de Somme. Havre de Carteret is the most northerly estuary along the western Cotentin coast. The estuary, situated at BarnevilleCarteret, is approximately 40 km south of the COGEMA-La Hague discharge pipeline (Figure 1), in the southwestern part of the coastal area where the highest activities of discharged radionuclides are observed (see above). A salt marsh core (BAR/3) was collected from an Atriplex portulacoides-dominated area of salt marsh adjacent to the main estuarine channel. Cored sediments were dominantly finegrained (40-60% of sediment 237Np > 239,240Pu. At the Baie de Somme site, although the 241Am profile is erratic due to the low activities observed (which are near detection limits), the (broad) 241Am peak occurs in the core at shallower depths than the initial increase in 137Cs (Figure 4). Again, this indicates that the main input of 241Am occurred after that of 137Cs, i.e., at a time period similar to that of the peak in 241Am observed at Barneville-Carteret. Pu Isotopic Ratio Data. At the Barneville-Carteret site, the source of the Pu was further constrained by determination of Pu isotopic composition. A COGEMA-La Hague, rather than weapons fallout, source for the Pu activity maximum in core BAR/3 is indicated by the ratio of 238Pu to 239,240Pu (Figure 6): 238Pu/239,240Pu increases steadily from 0.1 at -20 cm to 0.5 at the top of the core, which is significantly higher than values reported for atmospheric fallout (
